Do social links and connections have consequences for collective behavior? A central aspect of mass mobilization is diffusion of human, material and knowledge resources; while recent studies address its potential effects on collective action by examining networks and their properties, the lack of empirical data limits their insights to the realm of modeling. This study argues that particular configurations of opposition and government network typologies do affect mobilization outcomes, and advances this type of this research by formally examining four cases of mass mobilization [Serbia (1996-7, 2000) and Ukraine (2000-1, 2004)], through the use and analysis of specifically collected primary data. The paper introduces a networks perspective, framing the study in terms of individuals and their organizations–as parts of competing networks through which resources and behavior are communicated. It identifies two main rival political networks with key roles during the election campaign: the democratizing opposition that seeks to inform, recruit and mobilize the public while lowering protest thresholds, and the regime’s coercive apparatus, which tries to contain it. The study’s pluralist methodology includes data collection through extensive field research comprised of locating and interviewing key participants (2007-2013), the application of a modified snow-balling sampling method, and complemented archival research. The resulting data is converted into matrices and networks, with the metrics revealing their topology, emergent properties and performance dynamics vis a vis competing hypotheses, and specific network configurations. The ensuing analysis, including computer simulations, consists of a double, formal and empirical comparison-across different network types, and, across actual cases. Findings suggest that the combined effects of initial network structure and the evolution of the protest once electoral contestation is under way, affect diffusion processes and mobilization outcomes.